Research in Research in Psychology Psychology Psychology is defined as Psychology is defined as the scientific study of …… the scientific study of ……
Jan 10, 2016
Research in PsychologyResearch in Psychology
Psychology is defined as the Psychology is defined as the scientific study of ……scientific study of ……
Introducing ResearchIntroducing Research
1.1. What does it mean to say that someone has a What does it mean to say that someone has a scientific attitude?scientific attitude?
2.2. What do you think is the key distinction between a What do you think is the key distinction between a correlational study and an experiment?correlational study and an experiment?
3.3. Identify three types of correlational studies and Identify three types of correlational studies and provide some strengths and weaknesses of each.provide some strengths and weaknesses of each.
4.4. What role might confounding variables play in an What role might confounding variables play in an experimental study? Provide some examples.experimental study? Provide some examples.
5. What role do controls play in a research study? 5. What role do controls play in a research study? (Provide some examples.) (Provide some examples.)
Types of ResearchTypes of Research
Applied Research: Practical application..Applied Research: Practical application.. Clinical Research: Patient oriented..Clinical Research: Patient oriented.. Basic Research: Expand knowledge, no Basic Research: Expand knowledge, no
practical apppractical app
Why do we need research?Why do we need research?
The Dangers of Common SenseThe Dangers of Common Sense Hindsight BiasHindsight Bias or “I-knew-it-all-along” or “I-knew-it-all-along”
people tend to view events as more people tend to view events as more predictable than they really were.predictable than they really were.
IntuitionIntuition : : sensing without the use of rational processsensing without the use of rational process
False Consensus effectFalse Consensus effect – tendency to – tendency to overestimate others’ agreement with usoverestimate others’ agreement with us
OverconfidenceOverconfidence Confidence level: not a consistent predictor Confidence level: not a consistent predictor
of accuracy of accuracy
(We tend to remember our correct (We tend to remember our correct predictions and disregard our wrong ones!)predictions and disregard our wrong ones!)
Research StrategiesResearch Strategies
generate or refine
research and observations
lead to
hypothesis
theories
TheoryTheory an explanation that organizes and predicts an explanation that organizes and predicts
observationsobservations Hypothesis (Hypothesis (If…, thenIf…, then…)…)
a testable prediction often implied by a a testable prediction often implied by a theorytheory
Operational DefinitionsOperational Definitions the specific procedures or steps used in the the specific procedures or steps used in the
research studyresearch study
Scientific MethodScientific Method State the ProblemState the Problem Gather InformationGather Information Form a HypothesisForm a Hypothesis Test the HypothesisTest the Hypothesis Record and Analyze DataRecord and Analyze Data State the ConclusionState the Conclusion Repeat the WorkRepeat the Work
Types of ResearchTypes of Research
Correlational StudiesCorrelational Studies
Explores strength of a relationship Explores strength of a relationship between two or more variablesbetween two or more variables
Can make predictions, but cannot Can make predictions, but cannot explain cause and effectexplain cause and effect
Types of Correlational StudiesTypes of Correlational StudiesCase StudyCase Study In-depth study- one person / small In-depth study- one person / small
groupgroup StrengthStrength: provides valuable insight : provides valuable insight
on an issue / condition on an issue / condition WeaknessWeakness: findings can rarely be : findings can rarely be
generalized to a populationgeneralized to a population Some exceptions…(Phineas Gage)Some exceptions…(Phineas Gage)
Correlational Studies….Correlational Studies….Survey MethodSurvey Method research the research the self-reportedself-reported attitudes or attitudes or
behaviors of peoplebehaviors of people Questionnaire / Survey / interview...Questionnaire / Survey / interview... Strength: can be generalized to a Strength: can be generalized to a
larger population (if sample is random)larger population (if sample is random) Weakness: No cause and effectWeakness: No cause and effect
Wording of survey can introduce biasWording of survey can introduce bias
Correlational MethodsCorrelational Methods
Random SamplesRandom Samples Random Sample – when every person in Random Sample – when every person in
targeted population has equal chance of targeted population has equal chance of participating.participating.
The larger the sample size the more likely to The larger the sample size the more likely to represent the wholerepresent the whole
Ex. Computer generated list of every 5Ex. Computer generated list of every 5thth person. person.
Why not send survey the entire school?Why not send survey the entire school?
Correlational MethodsCorrelational Methods
Naturalistic ObservationNaturalistic Observation Observing subjects in natural Observing subjects in natural
habitats / no interaction between habitats / no interaction between researcher and subjectsresearcher and subjects
Strength: valuable insight into Strength: valuable insight into natural behaviornatural behavior
Weakness: no cause and effectWeakness: no cause and effect
Bias in Naturalistic ObservationBias in Naturalistic Observation Observer EffectObserver Effect: Behavior can change : Behavior can change
when being watched… (animals or when being watched… (animals or humans)humans)
Observer BiasObserver Bias: : When researchers notice what they want When researchers notice what they want
to see and ignore what doesn’t support to see and ignore what doesn’t support their theorytheir theory
Anthropomorphic FallacyAnthropomorphic Fallacy: Attributing : Attributing human thoughts, feelings, or motives to human thoughts, feelings, or motives to animals during research (especially when animals during research (especially when explaining behavior)explaining behavior)
Bias in research…Bias in research…
Experimenter EffectsExperimenter Effects: Changes in : Changes in behavior caused by the unintended behavior caused by the unintended influence of the experimenterinfluence of the experimenter
Self-Fulfilling ProphecySelf-Fulfilling Prophecy: When : When expectations affect outcome… expectations affect outcome… (teachers..)(teachers..)
Hawthorne EffectHawthorne Effect: subjects perception : subjects perception of experiment alters their performance...of experiment alters their performance...
Correlational Research Correlational Research Analyzing DataAnalyzing Data
Correlation Coefficient: measures the strength of a Correlation Coefficient: measures the strength of a relationship.relationship.
A Positive correlation: presence of one thing A Positive correlation: presence of one thing predicts the presence of the other thing. (2 sets of predicts the presence of the other thing. (2 sets of scores rise or fall together: +1=perfect correlation)scores rise or fall together: +1=perfect correlation)
A Negative correlation means the presence of one A Negative correlation means the presence of one thing predicts the absence of the other thing. (one thing predicts the absence of the other thing. (one goes up, the other down: ---1= perfect negative goes up, the other down: ---1= perfect negative
0 – no correlation0 – no correlation
Correlations: Statistical AnalysisCorrelations: Statistical Analysis
Scatterplot (aka scattergram, scatter diagram)Scatterplot (aka scattergram, scatter diagram) Graphed data that shows the strength of a Graphed data that shows the strength of a
correlationcorrelation The more the data forms a line, the The more the data forms a line, the
stronger the correlation, IOW:stronger the correlation, IOW: little scatter indicates high correlationlittle scatter indicates high correlation extensive scatter = little or no correlationextensive scatter = little or no correlation
the slope of the line suggests a positive or the slope of the line suggests a positive or negative correlationnegative correlation
Research MethodsResearch Methods
Perfect positivecorrelation (+1.00)
No relationship (0.00) Perfect negativecorrelation (-1.00)
Fig. 1.9 The relationship between
years of college completed and personal
income (hypothetical data).
Research StrategiesResearch Strategies
Scatter plot of Height and TemperamentScatter plot of Height and Temperament55 60 65 70 75 80 85
959085807570656055504540353025
Temperamentscores
Height in inches
Remember…Remember… Correlations do not prove causation.Correlations do not prove causation. Correlations only make predictions about Correlations only make predictions about
the relationship between two variablesthe relationship between two variables
Example: There is a positive correlation Example: There is a positive correlation between a child watching violent TV and between a child watching violent TV and that child displaying violent behavior. Why that child displaying violent behavior. Why can we not claim that violent TV causes can we not claim that violent TV causes violent behavior?violent behavior?
So…So…
Correlation indicates the Correlation indicates the possibilitypossibility of a cause-effect of a cause-effect relationship, but relationship, but it cannot proveit cannot prove causationcausation
Illusionary CorrelationIllusionary Correlation Illusionary Correlation – False assumption Illusionary Correlation – False assumption
of relationship between two things (tend to of relationship between two things (tend to recall instances that confirm our belief.)recall instances that confirm our belief.)
Examples:Examples: More babies born during a full moon, infertile More babies born during a full moon, infertile
couples who conceive after they adopt, couples who conceive after they adopt, premonition of a phone call followed by the premonition of a phone call followed by the phone callphone call
Remember…Remember…
Correlational methods can be Correlational methods can be used in case studies, surveys, used in case studies, surveys, and field experiences.and field experiences.
Experimentation: Cause and EffectExperimentation: Cause and Effect
Experiment: the clearest way to Experiment: the clearest way to isolate cause & effect isolate cause & effect
Experiments manipulate variables Experiments manipulate variables to test the resultsto test the results
Know the Difference…Know the Difference…
Correlational studies uncover Correlational studies uncover naturally occurring relationships.naturally occurring relationships.
experiments manipulates a experiments manipulates a setting to determine its effect.setting to determine its effect.
Key Experimental PrinciplesKey Experimental Principles
Experimental group – Group that Experimental group – Group that receives the treatmentreceives the treatment
Control group – Group that does Control group – Group that does not receive the treatment (placebo)not receive the treatment (placebo)
Each group must be from a Each group must be from a random random sample, and randomly assignedsample, and randomly assigned
Variables (Variables (Anything that can vary…)Anything that can vary…)
Independent Variable (IV) Independent Variable (IV) the variable that is manipulated the variable that is manipulated
(example, the drug)(example, the drug)
Dependent Variable (DV) Dependent Variable (DV) the variable that is measured (example: the variable that is measured (example:
behavior, or memory)behavior, or memory)
…….Variables.Variables
Determine the IV and DV in the Determine the IV and DV in the following hypothesis:following hypothesis:
IfIf you give a child sugar, you give a child sugar, thenthen the child’s activity level the child’s activity level increasesincreases..
ConfoundingConfounding Variables Variables
Inconsistencies between Inconsistencies between experimental and control groups experimental and control groups that can skew the results (bias!)that can skew the results (bias!) Examples: Time, place, frequency, Examples: Time, place, frequency,
amount etc.amount etc.
……Confounding VariablesConfounding Variables
If you give a child sugar, then the child’s If you give a child sugar, then the child’s activity level increasesactivity level increases..
What confounding principles might affect What confounding principles might affect this experiment?this experiment? Scheduled activities after the experimentScheduled activities after the experiment Location of observed behaviorLocation of observed behavior Peer group during experimentPeer group during experiment Size of peer group during experimentSize of peer group during experiment Amount and type of sugar ingestedAmount and type of sugar ingested Time of the experimentTime of the experiment Conditions prior to the experimentConditions prior to the experiment
Controls Controls (steps to avoid confounding (steps to avoid confounding variables)variables)
PlacebosPlacebos – “sugar pills” or any inert – “sugar pills” or any inert substance given to control group (used as substance given to control group (used as comparative basis against the experimental comparative basis against the experimental group)group)
Single blind procedureSingle blind procedure: Subjects don’t : Subjects don’t know which group they are in…know which group they are in…
Double-blind procedureDouble-blind procedure – neither subjects – neither subjects nor experimenter know if subjects are in the nor experimenter know if subjects are in the experimental group or the control groupexperimental group or the control group
Random sample and random assignmentRandom sample and random assignment
Remember…Remember…
Experiments aim to Experiments aim to manipulatemanipulate an an independentindependent variablevariable, , measuremeasure a a dependent variabledependent variable, and , and controlcontrol all other all other variablesvariables..
Good Research is…Good Research is…
Valid – when it measures what Valid – when it measures what the researcher set out to measure the researcher set out to measure and is accurateand is accurate
Reliable – when replication , with Reliable – when replication , with same operational definitions, same operational definitions, results in the same outcome.results in the same outcome.